diff --git a/.github/workflows/build-and-publish-pypi.yml b/.github/workflows/build-and-publish-pypi.yml
index 2731c97d..be8087de 100644
--- a/.github/workflows/build-and-publish-pypi.yml
+++ b/.github/workflows/build-and-publish-pypi.yml
@@ -20,6 +20,18 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v3
+      - uses: actions/setup-node@v4
+        with:
+          node-version: 18
+          cache: 'yarn'
+          cache-dependency-path: 'neetbox/frontend/yarn.lock'
+      - name: setup yarn
+        run: corepack enable
+        working-directory: neetbox/frontend
+      - run: yarn install --frozen-lockfile
+        working-directory: neetbox/frontend
+      - run: yarn build
+        working-directory: neetbox/frontend
       - name: Build and publish to pypi
         uses: JRubics/poetry-publish@v1.16
         with:
diff --git a/neetbox/integrations/resource.py b/neetbox/utils/resource.py
similarity index 97%
rename from neetbox/integrations/resource.py
rename to neetbox/utils/resource.py
index 0b82dc1e..1d3a054d 100644
--- a/neetbox/integrations/resource.py
+++ b/neetbox/utils/resource.py
@@ -1,276 +1,276 @@
-# -*- coding: utf-8 -*-
-#
-# Author: GavinGong aka VisualDust
-# URL:    https://gong.host
-# Date:   20230315
-
-import os
-import pathlib
-import signal
-import threading
-from concurrent.futures import ThreadPoolExecutor
-from functools import partial
-from random import random
-from threading import Event
-from typing import Dict, List, Union
-from urllib.request import urlopen
-
-import numpy as np
-from PIL import Image
-from rich.progress import (
-    BarColumn,
-    DownloadColumn,
-    Progress,
-    TaskID,
-    TextColumn,
-    TimeRemainingColumn,
-    TransferSpeedColumn,
-)
-
-from neetbox.integrations import engine
-from neetbox.logging import logger
-from neetbox.utils import pkg
-
-_loader_pool: Dict[str, "ResourceLoader"] = dict()  # all ResourceLoaders are stored here
-
-
-class ResourceLoader:
-    ready: bool = False  # stands for each scan
-    file_path_list: list = []
-    _initialized: bool = False  # stands for the first scan on creation
-
-    def __new__(
-        cls,
-        folder: str = ".",
-        file_types=["*"],
-        sub_dirs=True,
-        verbose=False,
-        force_rescan=False,
-        *args,
-        **kwargs,
-    ):
-        _id = folder + str(file_types) + "_R" if sub_dirs else ""
-        if _id in _loader_pool and not force_rescan:
-            logger.info(
-                "ResourceLoader with same path and same file type(s) already exists. Returning the old one."
-            )
-        else:
-            _loader_pool[_id] = super(ResourceLoader, cls).__new__(cls)
-        return _loader_pool[_id]
-
-    def __init__(
-        self,
-        folder: str = ".",
-        file_types=["*"],
-        sub_dirs=True,
-        async_scan=False,
-        verbose=False,
-        force_rescan=False,
-    ):
-        """ResourceLoader scans given file type(s) in given place(s)
-
-        Args:
-            folder (str): which folder to scan
-            file_types (str[]): file type(s) to include. For example, ['jpg','png']
-            sub_dirs (bool, optional): scan sub-folder(s)?. Defaults to True.
-            async_scan (bool, optional): run scan traks in a new thread. Defaults to False.
-            verbose (bool, optional): verbose output. Defaults to False.
-            force_rescan (bool, optional): rescan the folder even the same file type(s) was scanned here before. Default to False.
-        """
-        super().__init__()
-        self.path = os.path.abspath(folder)
-        self._file_types = file_types
-        self._scan_sub_dirs = sub_dirs
-        self._async_scan = async_scan
-        if not self.ready:
-            self._scan(verbose)
-
-    def _scan(self, verbose):
-        if not self.ready and self._initialized:
-            raise Exception("another scanning requested during the previous one.")
-        self.ready = False
-
-        def can_match(path: pathlib.Path):
-            if not path.is_file():
-                return False
-            for file_type in self._file_types:
-                if path.match("*." + file_type):
-                    return True
-            return False
-
-        def perform_scan():
-            glob_str = "**/*" if self._scan_sub_dirs else "*"
-            if not verbose:  # do not output
-                self.file_path_list = [
-                    str(path) for path in pathlib.Path(self.path).glob(glob_str) if can_match(path)
-                ]
-            else:
-                self.file_path_list = []
-                for path in pathlib.Path(self.path).glob(glob_str):
-                    if can_match(path):
-                        if verbose:
-                            logger.log(f"File match: {path}")
-                        self.file_path_list.append(path)
-            self.ready = True  # scan complete
-            if not self._initialized:
-                self._initialized = True
-            logger.ok(
-                f"Resource loader '{self.path}' ready with {'all' if '*' in self._file_types else len(self._file_types)} file types({len(self.file_path_list)} files)."
-            )
-
-        logger.log(
-            f"Scanning started at '{self.path}' for {'all' if '*' in self._file_types else len(self._file_types)} file types."
-        )
-        # call to scan
-        if self._async_scan:
-            threading.Thread(target=perform_scan).start()
-        else:
-            perform_scan()
-
-    def get_file_list(self):
-        if not self.ready:
-            raise Exception("not ready. scanning in process.")
-        return self.file_path_list.copy()
-
-    def __getitem__(self, index):
-        if not self.ready:
-            raise Exception("not ready. scanning in process.")
-        if type(index) is int:
-            return self.file_path_list[index]
-
-
-class ImagesLoader(ResourceLoader):
-    def __init__(
-        self,
-        folder,
-        file_types=["png", "jpg"],
-        sub_dirs=True,
-        async_scan=False,
-        verbose=False,
-    ):
-        pkg.is_installed("PIL", try_install_if_not="pillow")
-
-        super().__init__(folder, file_types, sub_dirs, async_scan, verbose)
-
-    def get_random_image(self):
-        rand_img_path = self.file_path_list[int(random() * len(self.file_path_list))]
-        image = Image.open(rand_img_path).convert("RGB")
-        return image
-
-    def get_random_images(self, howmany=1):
-        assert howmany < len(self.file_path_list)
-        rand_idx_begin = int(random() * (len(self.file_path_list) - howmany))
-        image_path_list = self[rand_idx_begin : rand_idx_begin + howmany]
-        image_list = []
-        for path in image_path_list:
-            image_list.append(Image.open(path).convert("RGB"))
-        return image_list
-
-    def get_random_image_as_numpy(self):
-        image = self.get_random_image()
-        return np.array(image)
-
-    def get_random_image_as_tensor(self, engine=engine.Torch):
-        assert engine in [engine.Torch]  # TODO support other engines
-        if engine == engine.Torch:
-            assert pkg.is_installed("torchvision")
-            import torchvision.transforms as T
-
-            tensor_transform = T.Compose(
-                [
-                    T.ToTensor(),
-                    T.Normalize(mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
-                ]
-            )
-            image = tensor_transform(self.get_random_image()).unsqueeze(0)  # To tensor of NCHW
-            return image
-
-    # TODO(VisualDust): to_dataset
-
-
-def download(
-    urls: Union[str, List[str], Dict[str, str]],
-    filenames: Union[str, List[str]] = None,
-    max_workers=4,
-):
-    """download both online and local files from urls
-
-    Example: download('')
-
-    Args:
-        urls (Union[str, List[str], Dict[str, str]]): single str to download from url; list of strs to download from urls; dict(filename:url) to download urls and rename them to filenames
-        filenames (Union[str, List[str]], optional): str to rename the downloaded file; list of strs to rename downloaded files; None means no rename. Defaults to None.
-        overwrite Bool: whether to skip exist files. Default to True.
-        retry (int, optional): retries when error occures. Defaults to 3.
-    """
-    progress = Progress(
-        TextColumn("[bold blue]{task.fields[filename]}", justify="right"),
-        BarColumn(bar_width=None),
-        "[progress.percentage]{task.percentage:>3.1f}%",
-        "•",
-        DownloadColumn(),
-        "•",
-        TransferSpeedColumn(),
-        "•",
-        TimeRemainingColumn(),
-    )
-
-    done_event = Event()
-
-    def handle_sigint(signum, frame):
-        done_event.set()
-
-    signal.signal(signal.SIGINT, handle_sigint)
-
-    def copy_url(task_id: TaskID, url: str, path: str) -> None:
-        """Copy data from a url to a local file."""
-        response = urlopen(url)
-        # This will break if the response doesn't contain content length
-        progress.update(task_id, total=int(response.info()["Content-length"]))
-        with open(path, "wb") as dest_file:
-            progress.start_task(task_id)
-            for data in iter(partial(response.read, 32768), b""):
-                dest_file.write(data)
-                progress.update(task_id, advance=len(data))
-                if done_event.is_set():
-                    return
-        logger.log(f"Downloaded {path}.")
-
-    assert type(urls) in [str, list, dict], "unkown format of url"
-
-    if type(urls) is str:
-        assert (
-            filenames is None
-            or type(filenames) is str
-            or (type(filenames) is list and len(filenames) == 1)
-        ), "num of urls and filenames mismatch"
-        urls = [
-            urls,
-        ]
-        filenames = (
-            [
-                filenames,
-            ]
-            if filenames
-            else None
-        )
-    if type(urls) is list:
-        if filenames is None:
-            filenames = [None] * len(urls)
-        assert type(filenames) is list and len(urls) == len(
-            filenames
-        ), "num of urls and filenames mismatch"
-
-        urls = {filenames[i]: urls[i] for i in range(len(urls))}
-
-    logger.log(f"Downloading {len(urls)} file(s)...")
-
-    with progress:
-        with ThreadPoolExecutor(max_workers=max_workers) as pool:
-            for fname, url in urls.items():
-                filename = url.split("/")[-1]
-                dest_path = os.path.join(fname)
-                task_id = progress.add_task("download", filename=filename, start=False)
-                pool.submit(copy_url, task_id, url, dest_path)
-
-    return urls
+# -*- coding: utf-8 -*-
+#
+# Author: GavinGong aka VisualDust
+# URL:    https://gong.host
+# Date:   20230315
+
+import os
+import pathlib
+import signal
+import threading
+from concurrent.futures import ThreadPoolExecutor
+from functools import partial
+from random import random
+from threading import Event
+from typing import Dict, List, Union
+from urllib.request import urlopen
+
+import numpy as np
+from PIL import Image
+from rich.progress import (
+    BarColumn,
+    DownloadColumn,
+    Progress,
+    TaskID,
+    TextColumn,
+    TimeRemainingColumn,
+    TransferSpeedColumn,
+)
+
+from neetbox.integrations import engine
+from neetbox.logging import logger
+from neetbox.utils import pkg
+
+_loader_pool: Dict[str, "ResourceLoader"] = dict()  # all ResourceLoaders are stored here
+
+
+class ResourceLoader:
+    ready: bool = False  # stands for each scan
+    file_path_list: list = []
+    _initialized: bool = False  # stands for the first scan on creation
+
+    def __new__(
+        cls,
+        folder: str = ".",
+        file_types=["*"],
+        sub_dirs=True,
+        verbose=False,
+        force_rescan=False,
+        *args,
+        **kwargs,
+    ):
+        _id = folder + str(file_types) + "_R" if sub_dirs else ""
+        if _id in _loader_pool and not force_rescan:
+            logger.info(
+                "ResourceLoader with same path and same file type(s) already exists. Returning the old one."
+            )
+        else:
+            _loader_pool[_id] = super(ResourceLoader, cls).__new__(cls)
+        return _loader_pool[_id]
+
+    def __init__(
+        self,
+        folder: str = ".",
+        file_types=["*"],
+        sub_dirs=True,
+        async_scan=False,
+        verbose=False,
+        force_rescan=False,
+    ):
+        """ResourceLoader scans given file type(s) in given place(s)
+
+        Args:
+            folder (str): which folder to scan
+            file_types (str[]): file type(s) to include. For example, ['jpg','png']
+            sub_dirs (bool, optional): scan sub-folder(s)?. Defaults to True.
+            async_scan (bool, optional): run scan traks in a new thread. Defaults to False.
+            verbose (bool, optional): verbose output. Defaults to False.
+            force_rescan (bool, optional): rescan the folder even the same file type(s) was scanned here before. Default to False.
+        """
+        super().__init__()
+        self.path = os.path.abspath(folder)
+        self._file_types = file_types
+        self._scan_sub_dirs = sub_dirs
+        self._async_scan = async_scan
+        if not self.ready:
+            self._scan(verbose)
+
+    def _scan(self, verbose):
+        if not self.ready and self._initialized:
+            raise Exception("another scanning requested during the previous one.")
+        self.ready = False
+
+        def can_match(path: pathlib.Path):
+            if not path.is_file():
+                return False
+            for file_type in self._file_types:
+                if path.match("*." + file_type):
+                    return True
+            return False
+
+        def perform_scan():
+            glob_str = "**/*" if self._scan_sub_dirs else "*"
+            if not verbose:  # do not output
+                self.file_path_list = [
+                    str(path) for path in pathlib.Path(self.path).glob(glob_str) if can_match(path)
+                ]
+            else:
+                self.file_path_list = []
+                for path in pathlib.Path(self.path).glob(glob_str):
+                    if can_match(path):
+                        if verbose:
+                            logger.log(f"File match: {path}")
+                        self.file_path_list.append(path)
+            self.ready = True  # scan complete
+            if not self._initialized:
+                self._initialized = True
+            logger.ok(
+                f"Resource loader '{self.path}' ready with {'all' if '*' in self._file_types else len(self._file_types)} file types({len(self.file_path_list)} files)."
+            )
+
+        logger.log(
+            f"Scanning started at '{self.path}' for {'all' if '*' in self._file_types else len(self._file_types)} file types."
+        )
+        # call to scan
+        if self._async_scan:
+            threading.Thread(target=perform_scan).start()
+        else:
+            perform_scan()
+
+    def get_file_list(self):
+        if not self.ready:
+            raise Exception("not ready. scanning in process.")
+        return self.file_path_list.copy()
+
+    def __getitem__(self, index):
+        if not self.ready:
+            raise Exception("not ready. scanning in process.")
+        if type(index) is int:
+            return self.file_path_list[index]
+
+
+class ImagesLoader(ResourceLoader):
+    def __init__(
+        self,
+        folder,
+        file_types=["png", "jpg"],
+        sub_dirs=True,
+        async_scan=False,
+        verbose=False,
+    ):
+        pkg.is_installed("PIL", try_install_if_not="pillow")
+
+        super().__init__(folder, file_types, sub_dirs, async_scan, verbose)
+
+    def get_random_image(self):
+        rand_img_path = self.file_path_list[int(random() * len(self.file_path_list))]
+        image = Image.open(rand_img_path).convert("RGB")
+        return image
+
+    def get_random_images(self, howmany=1):
+        assert howmany < len(self.file_path_list)
+        rand_idx_begin = int(random() * (len(self.file_path_list) - howmany))
+        image_path_list = self[rand_idx_begin : rand_idx_begin + howmany]
+        image_list = []
+        for path in image_path_list:
+            image_list.append(Image.open(path).convert("RGB"))
+        return image_list
+
+    def get_random_image_as_numpy(self):
+        image = self.get_random_image()
+        return np.array(image)
+
+    def get_random_image_as_tensor(self, engine=engine.Torch):
+        assert engine in [engine.Torch]  # TODO support other engines
+        if engine == engine.Torch:
+            assert pkg.is_installed("torchvision")
+            import torchvision.transforms as T
+
+            tensor_transform = T.Compose(
+                [
+                    T.ToTensor(),
+                    T.Normalize(mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
+                ]
+            )
+            image = tensor_transform(self.get_random_image()).unsqueeze(0)  # To tensor of NCHW
+            return image
+
+    # TODO(VisualDust): to_dataset
+
+
+def download(
+    urls: Union[str, List[str], Dict[str, str]],
+    filenames: Union[str, List[str]] = None,
+    max_workers=4,
+):
+    """download both online and local files from urls
+
+    Example: download('')
+
+    Args:
+        urls (Union[str, List[str], Dict[str, str]]): single str to download from url; list of strs to download from urls; dict(filename:url) to download urls and rename them to filenames
+        filenames (Union[str, List[str]], optional): str to rename the downloaded file; list of strs to rename downloaded files; None means no rename. Defaults to None.
+        overwrite Bool: whether to skip exist files. Default to True.
+        retry (int, optional): retries when error occures. Defaults to 3.
+    """
+    progress = Progress(
+        TextColumn("[bold blue]{task.fields[filename]}", justify="right"),
+        BarColumn(bar_width=None),
+        "[progress.percentage]{task.percentage:>3.1f}%",
+        "•",
+        DownloadColumn(),
+        "•",
+        TransferSpeedColumn(),
+        "•",
+        TimeRemainingColumn(),
+    )
+
+    done_event = Event()
+
+    def handle_sigint(signum, frame):
+        done_event.set()
+
+    signal.signal(signal.SIGINT, handle_sigint)
+
+    def copy_url(task_id: TaskID, url: str, path: str) -> None:
+        """Copy data from a url to a local file."""
+        response = urlopen(url)
+        # This will break if the response doesn't contain content length
+        progress.update(task_id, total=int(response.info()["Content-length"]))
+        with open(path, "wb") as dest_file:
+            progress.start_task(task_id)
+            for data in iter(partial(response.read, 32768), b""):
+                dest_file.write(data)
+                progress.update(task_id, advance=len(data))
+                if done_event.is_set():
+                    return
+        logger.log(f"Downloaded {path}.")
+
+    assert type(urls) in [str, list, dict], "unkown format of url"
+
+    if type(urls) is str:
+        assert (
+            filenames is None
+            or type(filenames) is str
+            or (type(filenames) is list and len(filenames) == 1)
+        ), "num of urls and filenames mismatch"
+        urls = [
+            urls,
+        ]
+        filenames = (
+            [
+                filenames,
+            ]
+            if filenames
+            else None
+        )
+    if type(urls) is list:
+        if filenames is None:
+            filenames = [None] * len(urls)
+        assert type(filenames) is list and len(urls) == len(
+            filenames
+        ), "num of urls and filenames mismatch"
+
+        urls = {filenames[i]: urls[i] for i in range(len(urls))}
+
+    logger.log(f"Downloading {len(urls)} file(s)...")
+
+    with progress:
+        with ThreadPoolExecutor(max_workers=max_workers) as pool:
+            for fname, url in urls.items():
+                filename = url.split("/")[-1]
+                dest_path = os.path.join(fname)
+                task_id = progress.add_task("download", filename=filename, start=False)
+                pool.submit(copy_url, task_id, url, dest_path)
+
+    return urls
diff --git a/pyproject.toml b/pyproject.toml
index 11bc3c4c..ca04f7eb 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "neetbox"
-version = "0.1.516"
+version = "0.2.0"
 description = "NEETBox contains useless CV code snippets."
 license = "MIT"
 authors = ["VisualDust <gavin@gong.host>"]
diff --git a/tests/test_integrations.py b/tests/test_integrations.py
index 25cfc7d4..2fe3299b 100644
--- a/tests/test_integrations.py
+++ b/tests/test_integrations.py
@@ -12,13 +12,16 @@
 #     for fname, furl in urls.items():
 #         os.remove(fname)
 
-# def test_resource_loader():
-#     from neetbox.integrations.resource import ResourceLoader
-#     import time
-#     print("testing resource loader")
-#     md_loader = ResourceLoader('./',['md'])
-#     py_loader_async = ResourceLoader('./', async_scan=True,verbose=True)
-#     while not py_loader_async.ready:
-#         print("waiting resource loader")
-#         time.sleep(1)
-#     py_loader_async.get_file_list()
+
+def test_resource_loader():
+    import time
+
+    from neetbox.utils.resource import ResourceLoader
+
+    print("testing resource loader")
+    md_loader = ResourceLoader("./", ["md"])
+    py_loader_async = ResourceLoader("./", async_scan=True, verbose=False)
+    while not py_loader_async.ready:
+        print("waiting resource loader")
+        time.sleep(1)
+    py_loader_async.get_file_list()